Jess Groopman, Industry Analyst with Altimeter Group at O’Reilly Solid
Last month, SmartBear was out in full form at O’Reilly Solid, a conference all about the Internet of Things (IoT) [this year]. We chatted with anyone and everyone who is building APIs to support their cool gadgets or back-end systems. What we found was…curious.
What is the Internet of Things to Everyone Else?
The concept behind the Internet of Things has been around for a while now, ATMs being some of the first enterprise, hardened, network-connected, managed devices for mainstream consumer use. So too with our mobile phones, these are not new concepts to network technicians or hardware geeks. But for the rest of us, we simply never imagined the extents that the “ubiquity of connectedness” would take all other industries, from biotech to automotive, personal care to agriculture, entertainment to custom manufacturing. The list is as long as our imaginations.
However, the IoT space is still very much a fledgling industry in many respects. The economics aren’t apparent; companies are still stuck under the “innovators bar,” finding something good enough to ship a minimum viable number of units. The safety and privacy considerations are staggeringly overlooked by both technology providers and end-users alike. Big businesses like IBM and HP are rushing to come up with strategies to retrofit existing solutions to fit the space or even create all new ones to support the new software production cultures of a post-agile tech world.
I had a chance to talk to a friend, Mark Boyd, long-time writer and traveling analyst for ProgrammableWeb and The New Stack about how the current state of the IoT industry seems to be so hyper-focused on innovation.
“It seems for these folks [focused on IoT at O’Reilly Solid] to be very much about getting to the point where they’re shipping 10,000 units…getting the product out there to an audience”, says Mark. “It’s not so much about solving the problem of integration or how to make that hardware work with other hardware.”
Innovation IS Integration, and Must Happen at the Same Time
From a broader development and ecosystem perspective, a “worry about integration later” approach could be a very costly short-sighted blunder, even for businesses just getting off the ground. Much of the protocols and hardware platforms have still remained proprietary, representing a barrier of entry for mainstream developers. From the very beginning, selecting open standards that are safe and reliable is a minimum viable design decision that should be made early on, specifically to foster rapid acceptance by developers and encourage partnerships to larger players in the field too.
The same goes for aspects of privacy and data ownership over the information being produced from IoT devices. It’s just too easy to plug something in and connect it up…without considering the implications of doing just that. I asked Mark about this too, and he continued:
“I don’t think the data privacy or data ownership considerations are thought through early enough in the conversation, especially around the rights of the end user to be able to access their own data or confidence in where it’s stored. “ Interesting, but he wasn’t done there. “On the other side, what gives me a bit of relief is, McKinsey Global Institute presented their IoT report here yesterday, fascinating. One of the things Michael Chui talked about was how little of the data exhaust is being looked at. So there is a whole lot of IoT big data, but because humans are lazy, I think the nefarious uses are something that no one has been looking at. I think there is the potential for problems, but at the moment the main use of IoT data is really around identifying alerts in the system and then for communicating real time data flow…but then all the data exhaust about given behavior that’s been collected, none of that’s being used.”
While Mark jokes, he also expressed hope that this breathing space we have before the data exhaust gets used allows for more time to openly discuss and educate around data rights. After our interview, Mark shared that one possible ideal vision would be for all of a user’s data to be stored by a third party data broker and that it could then be sold back to the device makers (or consent given to the device maker to collect it directly and receive, for example, a discounted subscription fee) or on to other data wranglers. In this way, an individual’s data would be akin to their currency and the storage of personal data akin to a bank account. Mark shared that one of the benefits of this sort of approach would be that it intrinsically symbolizes the value of an individual’s data so that each of us can consider who we share our data with or what we “spend” it on, as it were.
You Might Not Be Looking at Your Own Data, but Who Is?
An immediate lack of interest over exhaust data makes sense; we’re already inundated with oceans of data all day long, even if you aren’t in a high-tech industry. But there’s little cost to storing that data long-term, and all it takes is a fast, very small bit of coagulation in the IoT technology space (specifically around protocols) to realize the potential of that data. Simply map-reduce, aggregate, and extract intelligence. It’s not hard, and insurance companies already use your Fitbit data to set pricing and fee structures.
Riding on the heels of her most excellent presentation at APIdays San Francisco the week before O’Reilly Solid, Jessica Groopman of the Altimeter Group recently released a research report “Consumer Perceptions of Privacy in the Internet of Things” which indicates that one of the foremost barrier facing IoT is consumer concern over use and sharing of connected device data.
Mark finishes with, “And here’s again where APIs would enable 3rd parties to be able to start innovating and creating new products from that data exhaust”, a topic near and dear to our hearts here at SmartBear: the connected world through the lens of quality. APIs, apps, solutions…are nothing if they’re not accurate, reliable, and safe. No house stands the test of time unless the foundation is solid.
What’s Next for the Internet of Things?
Two words: big intelligence. There is a huge opportunity to extract meaningful opportunities from IoT data exhaust once it is collected and normalized. That places a huge amount of responsibility over big platform companies to become mass repositories of IoT data, but especially responsible are the IoT vendors who must make a choice over which platform to use.
Fortunately, the IoT industry not only benefits from years of traditional field experience (such as in ATMs, cell services, medical and industrial equipment) but also from recent mass adoption of API design and incremental architecture patterns like Microservices. To Mark’s point, true innovation in the IoT space will only really explode once the immediate hardware hype gentrifies to a point where open standards allow throngs of developers to iterate, poke, and prod at the big data, not just twiddle with the devices.
Once IoT data is in a platform, more than just developers can realize its full potential. Data analysts, business decision experts, and policy makers will all have very critical jobs in the IoT space. In order for a technology to become a brand new industry, it must by definition include people from all areas of business, not just hardware and software geeks (like us). IoT is rapidly approaching that point.
Written by Paul Bruce